P
US7843162B2ActiveUtilityPatentIndex 83

Current regulator and current control method and system for AC motors

Assignee: GM GLOBAL TECH OPERATIONS INCPriority: Mar 13, 2008Filed: Mar 13, 2008Granted: Nov 30, 2010
Est. expiryMar 13, 2028(~1.7 yrs left)· nominal 20-yr term from priority
Inventors:BAE BONHOPATEL NITINKUMAR RSCHULZ STEVEN ENAGASHIMA JAMES MYIM JUNG SIKSUL SEUNG-KI
H02P 21/22
83
PatentIndex Score
13
Cited by
3
References
20
Claims

Abstract

Methods and systems are provided for controlling an AC motor via an inverter. The method includes determining a delay-compensated offset based on a synchronous frame current, producing a current error based on a synchronous frame current and a commanded current, producing a voltage error based on an anti-windup offset and the current error, producing a commanded voltage based on the delay-compensated offset and the voltage error, and providing the inverter with the commanded voltage.

Claims

exact text as granted — not AI-modified
1. A method for controlling an alternating current (AC) motor via an inverter, the AC motor operable based on a commanded current to produce a stator current, the method comprising the steps of:
 determining a delay-compensated offset based on a synchronous frame current, the synchronous frame current derived from the stator current; 
 producing a current error based on the synchronous frame current and the commanded current; 
 producing a voltage error based on an anti-windup offset and the current error, the anti-windup offset associated with a pre-determined voltage limit of the inverter; 
 producing a commanded voltage based on the delay-compensated offset and the voltage error; and 
 providing the inverter with the commanded voltage. 
 
     
     
       2. A method according to  claim 1 , wherein the inverter has a sampling period and a damping resistance, and wherein the step of determining a delay-compensated offset comprises:
 predicting a delay-compensated current based on the sampling period and the synchronous frame current; and 
 determining the delay-compensated offset based on the delay-compensated current and the damping resistance. 
 
     
     
       3. A method according to  claim 2 , wherein the commanded current comprises a d-axis commanded current and a q-axis commanded current, wherein the delay-compensated current comprises a d-axis delay-compensated current and a q-axis delay-compensated current, wherein the synchronous frame current comprises a d-axis synchronous frame current and a q-axis synchronous frame current, and wherein the step of predicting comprises:
 calculating the d-axis delay-compensated current based on the d-axis synchronous frame current and a first sampling factor, the d-axis synchronous frame current based on the d-axis commanded current, the first sampling factor based on a q-axis stator flux of the AC motor and a d-axis stator inductance of the AC motor; and 
 calculating the q-axis delay-compensated current based on the q-axis synchronous frame current and a second sampling factor, the q-axis synchronous frame current based on the q-axis commanded current, the second sampling factor based on a d-axis stator flux of the AC motor and a q-axis stator inductance of the AC motor. 
 
     
     
       4. A method according to  claim 3 , wherein the q-axis stator flux, the d-axis stator flux, the d-axis stator inductance, and the q-axis stator inductance are each a function of the d-axis commanded current and the q-axis commanded current. 
     
     
       5. A method according to  claim 3 , wherein the step of calculating the d-axis delay-compensated current (i r   ds ′[k]) comprises calculating 
       
         
           
             
               
                 
                   
                     i 
                     ds 
                     
                       r 
                       ′ 
                     
                   
                   ⁡ 
                   
                     [ 
                     k 
                     ] 
                   
                 
                 = 
                 
                   
                     
                       i 
                       ds 
                       r 
                     
                     ⁡ 
                     
                       [ 
                       k 
                       ] 
                     
                   
                   + 
                   
                     
                       T 
                       samp 
                     
                     ⁢ 
                     
                       
                         
                           
                             v 
                             ds 
                             
                               r 
                               * 
                             
                           
                           ⁡ 
                           
                             [ 
                             
                               k 
                               - 
                               1 
                             
                             ] 
                           
                         
                         - 
                         
                           
                             R 
                             s 
                           
                           ⁢ 
                           
                             
                               i 
                               ds 
                               
                                 r 
                                 * 
                               
                             
                             ⁡ 
                             
                               [ 
                               k 
                               ] 
                             
                           
                         
                         + 
                         
                           
                             ω 
                             r 
                           
                           ⁢ 
                           
                             
                               λ 
                               q 
                             
                             ⁡ 
                             
                               ( 
                               
                                 
                                   
                                     i 
                                     ds 
                                     
                                       r 
                                       * 
                                     
                                   
                                   ⁡ 
                                   
                                     [ 
                                     k 
                                     ] 
                                   
                                 
                                 , 
                                 
                                   
                                     i 
                                     qs 
                                     
                                       r 
                                       * 
                                     
                                   
                                   ⁡ 
                                   
                                     [ 
                                     k 
                                     ] 
                                   
                                 
                               
                               ) 
                             
                           
                         
                       
                       
                         
                           L 
                           d 
                         
                         ⁡ 
                         
                           ( 
                           
                             
                               
                                 i 
                                 ds 
                                 
                                   r 
                                   * 
                                 
                               
                               ⁡ 
                               
                                 [ 
                                 k 
                                 ] 
                               
                             
                             , 
                             
                               
                                 i 
                                 qs 
                                 
                                   r 
                                   * 
                                 
                               
                               ⁡ 
                               
                                 [ 
                                 k 
                                 ] 
                               
                             
                           
                           ) 
                         
                       
                     
                   
                 
               
               ; 
               and 
             
           
         
         wherein the step of calculating the q-axis delay-compensated current (i r   qs ′[k]) comprises calculating 
       
       
         
           
             
               
                 
                   
                     i 
                     qs 
                     
                       r 
                       ′ 
                     
                   
                   ⁡ 
                   
                     [ 
                     k 
                     ] 
                   
                 
                 = 
                 
                   
                     
                       i 
                       qs 
                       r 
                     
                     ⁡ 
                     
                       [ 
                       k 
                       ] 
                     
                   
                   + 
                   
                     
                       T 
                       samp 
                     
                     ⁢ 
                     
                       
                         
                           
                             v 
                             qs 
                             
                               i 
                               * 
                             
                           
                           ⁡ 
                           
                             [ 
                             
                               k 
                               - 
                               1 
                             
                             ] 
                           
                         
                         - 
                         
                           
                             R 
                             s 
                           
                           ⁢ 
                           
                             
                               i 
                               qs 
                               
                                 r 
                                 * 
                               
                             
                             ⁡ 
                             
                               [ 
                               k 
                               ] 
                             
                           
                         
                         - 
                         
                           
                             ω 
                             r 
                           
                           ⁢ 
                           
                             
                               λ 
                               d 
                             
                             ⁡ 
                             
                               ( 
                               
                                 
                                   
                                     i 
                                     ds 
                                     
                                       r 
                                       * 
                                     
                                   
                                   ⁡ 
                                   
                                     [ 
                                     k 
                                     ] 
                                   
                                 
                                 , 
                                 
                                   
                                     i 
                                     qs 
                                     
                                       r 
                                       * 
                                     
                                   
                                   ⁡ 
                                   
                                     [ 
                                     k 
                                     ] 
                                   
                                 
                               
                               ) 
                             
                           
                         
                       
                       
                         
                           L 
                           q 
                         
                         ⁡ 
                         
                           ( 
                           
                             
                               
                                 i 
                                 ds 
                                 
                                   r 
                                   * 
                                 
                               
                               ⁡ 
                               
                                 [ 
                                 k 
                                 ] 
                               
                             
                             , 
                             
                               
                                 i 
                                 qs 
                                 
                                   r 
                                   * 
                                 
                               
                               ⁡ 
                               
                                 [ 
                                 k 
                                 ] 
                               
                             
                           
                           ) 
                         
                       
                     
                   
                 
               
               , 
             
           
         
         wherein i r   ds [k] is the d-axis synchronous frame, i r   qs [k] is the q-axis synchronous frame current produced, T samp  is a sampling period, v r*   ds [k−1] is a d-axis commanded voltage based on i r*   ds [k−1], v r*   qs [k−1] is a q-axis commanded voltage based on i r*   qs [k−1], R s  is an equivalent stator resistance of the AC motor, ω r  is a synchronous frequency of the AC motor, λ d (i r*   ds [k], i r*   qs [k]) is a d-axis stator flux of the AC motor, λ q (i r*   ds [k], i r*   qs [k]) is a q-axis stator flux of the AC motor, L d (i r*   ds [k], i r*   qs [k]) is a d-axis stator inductance of the AC motor, L q (i r*   ds [k], i r*   qs [k]) is a q-axis stator inductance of the AC motor, and, k represents an integer. 
       
     
     
       6. A method according to  claim 1 , wherein the step of producing a voltage error comprises:
 applying a pre-determined gain to the clamped voltage due to the pre-determined voltage limit of the inverter to produce the anti-windup offset; and 
 subtracting the anti-windup offset from the current error. 
 
     
     
       7. A method according to  claim 1 , wherein the step of producing the commanded voltage comprises:
 combining the voltage error with a feedforward term, a proportional term, and the delay-compensated offset to produce the commanded voltage, the feedforward term based on a d-axis voltage, the proportional term based on the current error. 
 
     
     
       8. A method according to  claim 1 , wherein the commanded current comprises first and second commanded currents, wherein the synchronous frame current comprises first and second synchronous frame currents, wherein the current error comprises first and second current errors, the first current error based on the first commanded current and the first synchronous frame current, the second current error based on the second commanded current and the second synchronous frame current, and wherein the step of producing a voltage error comprises:
 combining the first current error with the anti-windup offset to produce a first term; 
 applying a gain to the first term to produce a second term; 
 combining the second term with a cross-coupling term to produce a third term, the cross-coupling term based on the second current error; and 
 integrating the third term to produce the voltage error. 
 
     
     
       9. A method according to  claim 8 , further comprising low-pass filtering the second current error to produce the cross-coupling term. 
     
     
       10. A current regulator for controlling an alternating current (AC) motor via a commanded voltage in response to a commanded current, the AC motor operable to produce a stator current at a fundamental frequency based on the commanded voltage, the current regulator comprising:
 an input configured to receive the commanded current and a synchronous frame current, the synchronous frame current based on the stator current; 
 an output configured to provide the commanded voltage; and 
 a processor coupled to the input and the output, the processor configured to: 
 sample the commanded current and the synchronous frame current at a sampling frequency; and 
 produce the commanded voltage based on the commanded current and the synchronous frame current, the commanded voltage being compensated for a delay and an anti-windup offset, the anti-windup offset associated with a pre-determined voltage limit. 
 
     
     
       11. A current regulator according to  claim 10 , wherein the processor is further configured to:
 predict a delay-compensated current based on the sampling frequency and the synchronous frame current; 
 determine a delay-compensated offset based on the delay-compensated current and a pre-determined damping resistance; and 
 compensate the commanded voltage for the delay with the delay-compensated offset. 
 
     
     
       12. A current regulator according to  claim 11 , wherein the commanded current comprises a d-axis commanded current and a q-axis commanded current, wherein the delay-compensated current comprises a d-axis delay-compensated current and a q-axis delay-compensated current, wherein the synchronous frame current comprises a d-axis current and a q-axis current, and wherein the processor is further configured to:
 calculate the d-axis delay-compensated current based on the d-axis current and a first sampling factor, the d-axis current based on the d-axis commanded current, the first sampling factor based on a q-axis stator flux of the AC motor and a d-axis stator inductance of the AC motor; and 
 calculate the q-axis delay-compensated current based on the q-axis current and a second sampling factor, the q-axis current based on the q-axis commanded current, the second sampling factor based on a d-axis stator flux of the AC motor and a q-axis stator inductance of the AC motor. 
 
     
     
       13. A current regulator according to  claim 12 , wherein the processor is further configured to:
 calculate the d-axis delay-compensated current (i r   ds ′[k]) from 
 
       
         
           
             
               
                 
                   
                     r 
                     qs 
                     
                       r 
                       ′ 
                     
                   
                   ⁡ 
                   
                     [ 
                     k 
                     ] 
                   
                 
                 = 
                 
                   
                     
                       i 
                       qs 
                       r 
                     
                     ⁡ 
                     
                       [ 
                       k 
                       ] 
                     
                   
                   + 
                   
                     
                       T 
                       samp 
                     
                     ⁢ 
                     
                       
                         
                           
                             v 
                             qs 
                             
                               r 
                               * 
                             
                           
                           ⁡ 
                           
                             [ 
                             
                               k 
                               - 
                               1 
                             
                             ] 
                           
                         
                         - 
                         
                           
                             R 
                             s 
                           
                           ⁢ 
                           
                             
                               i 
                               ds 
                               
                                 r 
                                 * 
                               
                             
                             ⁡ 
                             
                               [ 
                               k 
                               ] 
                             
                           
                         
                         + 
                         
                           
                             ω 
                             r 
                           
                           ⁢ 
                           
                             
                               λ 
                               q 
                             
                             ⁡ 
                             
                               ( 
                               
                                 
                                   
                                     i 
                                     ds 
                                     
                                       r 
                                       * 
                                     
                                   
                                   ⁡ 
                                   
                                     [ 
                                     k 
                                     ] 
                                   
                                 
                                 , 
                                 
                                   
                                     i 
                                     qs 
                                     
                                       r 
                                       * 
                                     
                                   
                                   ⁡ 
                                   
                                     [ 
                                     k 
                                     ] 
                                   
                                 
                               
                               ) 
                             
                           
                         
                       
                       
                         
                           L 
                           d 
                         
                         ⁡ 
                         
                           ( 
                           
                             
                               
                                 i 
                                 ds 
                                 
                                   r 
                                   * 
                                 
                               
                               ⁡ 
                               
                                 [ 
                                 k 
                                 ] 
                               
                             
                             , 
                             
                               
                                 i 
                                 qs 
                                 
                                   r 
                                   * 
                                 
                               
                               ⁡ 
                               
                                 [ 
                                 k 
                                 ] 
                               
                             
                           
                           ) 
                         
                       
                     
                   
                 
               
               ; 
               and 
             
           
         
         calculate the q-axis delay-compensated current (i r   qs ′[k]) from 
       
       
         
           
             
               
                 
                   
                     r 
                     qs 
                     
                       r 
                       ′ 
                     
                   
                   ⁡ 
                   
                     [ 
                     k 
                     ] 
                   
                 
                 = 
                 
                   
                     
                       i 
                       qs 
                       r 
                     
                     ⁡ 
                     
                       [ 
                       k 
                       ] 
                     
                   
                   + 
                   
                     
                       T 
                       samp 
                     
                     ⁢ 
                     
                       
                         
                           
                             v 
                             qs 
                             
                               r 
                               * 
                             
                           
                           ⁡ 
                           
                             [ 
                             
                               k 
                               - 
                               1 
                             
                             ] 
                           
                         
                         - 
                         
                           
                             R 
                             s 
                           
                           ⁢ 
                           
                             
                               i 
                               qs 
                               
                                 r 
                                 * 
                               
                             
                             ⁡ 
                             
                               [ 
                               k 
                               ] 
                             
                           
                         
                         - 
                         
                           
                             ω 
                             r 
                           
                           ⁢ 
                           
                             
                               λ 
                               d 
                             
                             ⁡ 
                             
                               ( 
                               
                                 
                                   
                                     i 
                                     ds 
                                     
                                       r 
                                       * 
                                     
                                   
                                   ⁡ 
                                   
                                     [ 
                                     k 
                                     ] 
                                   
                                 
                                 , 
                                 
                                   
                                     i 
                                     qs 
                                     
                                       r 
                                       * 
                                     
                                   
                                   ⁡ 
                                   
                                     [ 
                                     k 
                                     ] 
                                   
                                 
                               
                               ) 
                             
                           
                         
                       
                       
                         
                           L 
                           q 
                         
                         ⁡ 
                         
                           ( 
                           
                             
                               
                                 i 
                                 ds 
                                 
                                   r 
                                   * 
                                 
                               
                               ⁡ 
                               
                                 [ 
                                 k 
                                 ] 
                               
                             
                             , 
                             
                               
                                 i 
                                 qs 
                                 
                                   r 
                                   * 
                                 
                               
                               ⁡ 
                               
                                 [ 
                                 k 
                                 ] 
                               
                             
                           
                           ) 
                         
                       
                     
                   
                 
               
               , 
             
           
         
         wherein i r   ds [k] is the d-axis current, i r   qs [k] is the q-axis current, T samp  is a sampling period based on the sampling frequency, v r*   ds [k−1] is a d-axis commanded voltage based on i r*   ds [k−1], v r*   qs [k−1] is a q-axis commanded voltage based on i r*   qs [k−1], R s  is an equivalent stator resistance of the AC motor, ω r  is a synchronous frequency of the AC motor, λ d (i r*   ds [k], i r   qs [k]) is a d-axis stator flux of the AC motor, λ q (i r*   ds [k], i r   qs [k]) is a q-axis stator flux of the AC motor, L d (i r*   ds [k], i r   qs [k]) is a d-axis stator inductance of the AC motor, L q (i r*   ds [k], i r   qs [k]) is a q-axis stator inductance of the AC motor, and, k represents an integer. 
       
     
     
       14. A current regulator according to  claim 11 , wherein the processor is further configured to:
 produce a current error based on the synchronous frame current and the commanded current; 
 apply a pre-determined gain to the clamped voltage due to the pre-determined voltage limit to produce the anti-windup offset; 
 subtract the anti-windup offset from the current error to produce a voltage error; and 
 produce the commanded voltage based on the delay-compensated offset and the voltage error. 
 
     
     
       15. A current regulator according to  claim 14 , wherein the commanded current comprises first and second commanded currents, wherein the synchronous frame current comprises first and second synchronous frame currents, wherein the current error comprises first and second current errors, the first current error based on the first commanded current and the first synchronous frame current, the second current error based on the second commanded current and the second synchronous frame current, and wherein the processor is further configured to:
 combine the first current error with the anti-windup offset to produce a first term; 
 apply a gain to the first term to produce a second term; 
 combine the second term with a cross-coupling term to produce a third term, the cross-coupling term based on the second current error; and 
 integrate the third term to produce the voltage error. 
 
     
     
       16. A current regulator according to  claim 15 , wherein the processor comprises a low-pass filter for filtering the second current error to produce the cross-coupling term. 
     
     
       17. A system for controlling an alternating current (AC) motor, the system comprising:
 an inverter having an output configured to couple to the AC motor, the inverter configured to produce an AC voltage based on a commanded voltage and further configured to drive the AC motor at a fundamental frequency with the AC voltage, the AC motor producing a stator current based on the AC voltage; and 
 a controller coupled to the inverter, the controller having a sampling frequency and configured to produce a commanded voltage based on a commanded current and a synchronous frame current, the commanded voltage being compensated for a delay and an anti-windup offset, the synchronous frame current based on the stator current, the anti-windup offset associated with a pre-determined voltage limit. 
 
     
     
       18. A system according to  claim 17 , wherein the inverter is an asynchronous pulse width modulation (PWM) inverter having an input coupled to the controller. 
     
     
       19. A system according to  claim 18 , wherein the AC motor has a pole count greater than about six. 
     
     
       20. A system according to  claim 17 , wherein the ratio of the sampling frequency to a fundamental frequency is less than about twenty-one.

Cited by (0)

No later patents cite this yet.

References (0)

No backward citations on record.